![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081013073347im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
Statistical Modeling and Analyses Plans
Research Programs
Modeling
- To develop tools to help EPA and state air-quality managers determine whether there have been significant changes in regional ozone concentrations and extreme values (daily maximum 1-hour and 8-hour) levels in the eastern U.S. since the implementation of NOx reductions stemming from the NOx SIP call.
- If such changes are discovered, to develop statistical methods to determine to what extent the reductions in emissions may have played a role in the changes and to determine what role future reductions may be expected to play, considering natural variations in air quality.
Analyses
NOx Accountability
- Assessing ozone levels over time at a specific location or region:
- Assessing ozone levels over a reiona
A time series models for extreme values will be developed, based on the observational data, which allows us to separate the trends from the noise components. This will enable us to assess the trends in ozone levels over time and to assess how well the model fits in statistical terms (confidence limits, margins of error, etc.). The fitted model would help us determine whether there have been increases or decreases in ozone levels over time, whether seasonal trends have changed, and whether any changes seem to be a significant in relation to the amount of variability (noise) seen in the data.
For a given time (or time period), a mdoel will be developed to describe the spatial distribution of ozone throughout the eastern U.S. Using this mdoel, we would be able to assess the overall pattern of ozone distribution and to assess how well the model fits in statistical terms. Such a spatial anaysis is important because it allows use ask whether ozone levels have changed in one region (but not others) or whether the ozone problem is shifting from one region to anothe region, for example.